Analysis Source of Ignition and Fuel for the Fire Normally, for liquid fuel to ignite, it must be above its flash point for vapours to produce an ignitable mixture. In this instance, two exhaust manifolds with temperatures of approximately 1150F (620C) were present within 60cm of the failed fitting/connections. The flash point of the fuel oil supplying the generator was determined to be 68C. In the case of fuel released under pressure, as in this instance, ignition can take place at a temperature below the flash point of the fuel, provided that a source of ignition above the flash point of the fuel is present. Given the fire burn patterns and the fuel source for ignition, the exhaust manifolds are considered to be the source of ignition for the fire. The partial separation of the fuel fitting from the filter housing in conjunction with a fuel pressure of 4.0bar would produce a readily ignitable atomized spray. Following the occurrence, when the fuel oil system had been pressurized, the O-ring elbow of the failed fitting started to spray fuel oil at 1.5bar, i.e., some 37.5percent of the normal fuel oil pressure. Given the position of the exhaust manifolds, the fuel oil released from the failed fitting/connection on the secondary fuel oil filter housing is considered to be the primary source of fuel for starting the fire. Once the fire was established, a flexible fuel hose, which connected a composite signal line from the primary fuel oil filter to the low fuel oil pressure switch, burned through and failed. This resulted in a secondary source of fuel to feed and maintain the fire. Impact of the Deactivation of the Fuel Alarm and Monitoring System It is unlikely that the low fuel oil pressure alarm, even if armed, would be triggered in a timely manner prior to the onset of the fire, given: the difference in value between normal and alarm fuel oil pressures; that the fuel oil supply to consumption ratio is greater than four to one; and the relatively small size of the fuel fitting/connection leak. Cause of Blackout Prior to re-establishing main and auxiliary power onboard the vessel, an inspection of the quick-closing fuel valves by the chief engineer revealed that valves to both engine rooms were in the tripped position. This resulted in fuel starvation for all machinery in the port and starboard engine rooms and the subsequent blackout. The blackout occurred at the time the starboard engine room fuel trip had been activated. This would suggest that accidental activation of the port engine room fuel trip occurred around this time and would account for the blackout. An emergency generator is designed to start up and put itself on load in the event of a blackout. Given that time was of the essence, the chief engineer intervened and completed the process manually. Engine Room Exhaust Ventilation When originally fitted at the shipyard, the dampers on the engine room exhaust ventilation were designed to be actuated remotely through the use of an electric solenoid. Ideally, a system such as this should be designed to fail in the closed position (fail close), i.e., any loss of electrical power results in the damper closing. During the subsequent fire damage repairs, the ventilation dampers were observed to have failed in the open position (fail open), i.e., they required power to close. It is most likely that, immediately after the onset of the fire, the electrical wiring which supplied power to the solenoid burned through and failed. With the loss of the electrical circuit to the solenoid, the engine room exhaust ventilation dampers could not be operated and, therefore, remained open. Emergency Response The vessel is equipped with two separate engine rooms and only the starboard engine room fuel shut-off valve had been activated. As the electrical load had been transferred to three main generators in the port engine room, it would indicate that fuel was available to the port engine room. Soon after activating the starboard engine room fuel shut-off valve, the vessel suffered a total loss of all main and auxiliary electrical power. Given the small complement on board and, as time was of the essence, priority was placed on activities associated with extinguishing the fire and the cause of the blackout was not investigated. After the fire was extinguished, and during the subsequent start-up of the main generators in the port engine room, it was discovered the blackout was due to the inadvertent activation of the port engine room fuel shut-off valve. For this to have occurred, either the petcock for the port engine room shut-off valve may not have been fully closed or sufficient hydraulic pressure may have built up in the port emergency fuel shut-off system as a result of the fire, to activate the valve. Once the fire had been extinguished, power to the port engine room was restored and the vessel returned to port under her own power. The speed and efficacy in which the crew assembled and handled the emergency is directly attributed to their emergency response drills and exercise regime. Carried out on a weekly basis and at a frequency four times that required by regulation, the familiarity of the crew with the vessel and its emergency protocols and equipment was a significant factor in the timely and effective manner in which the fire was fought. Stack Fire The exact cause of the No.6 main generator stack fire could not be determined. One possible explanation is that, when the remote fuel trips were activated, the resulting loss in fuel to the running generator caused it to slow down. The governor, sensing the generator was now operating below rated rpm, went to a full fuel position. The generator, effectively starved of fuel, continued to slow down, resulting in incomplete combustion of the fuel being supplied to the cylinders. The incomplete combustion could have resulted in the dumping of raw fuel into the generator uptake, which ultimately flashed off due to the heat of the fire. The source of fuel for the fire was the failed fuel oil connection on the secondary fuel oil filter. The cause for the fuel oil fitting/connection failure was the backing off of the elbow locknut and the resulting relative motion between the two components. The sources of ignition for the fire were the generator exhaust manifolds. A secondary source of fuel for the fire was the burnt fuel oil pressure signal line from the primary fuel oil filter to the low fuel oil pressure alarm switch.Findings as to Causes and Contributing Factors The source of fuel for the fire was the failed fuel oil connection on the secondary fuel oil filter. The cause for the fuel oil fitting/connection failure was the backing off of the elbow locknut and the resulting relative motion between the two components. The sources of ignition for the fire were the generator exhaust manifolds. A secondary source of fuel for the fire was the burnt fuel oil pressure signal line from the primary fuel oil filter to the low fuel oil pressure alarm switch. The use of a combination of straight and tapered threads in a pressurized system has a potential to generate leaks under certain circumstances. The engine room exhaust ventilation dampers were installed such that an electrical source of power was required to activate (close) them.Findings as to Risk The use of a combination of straight and tapered threads in a pressurized system has a potential to generate leaks under certain circumstances. The engine room exhaust ventilation dampers were installed such that an electrical source of power was required to activate (close) them. The speed and efficiency with which the emergency was handled was directly attributable to the frequency of the emergency preparedness drills carried out onboard the vessel. The inadvertent closing of the port engine room fuel shut-off valve resulted in a blackout of the vessel. The exact cause of the stack fire on No.6 main generator could not be determined.Other Findings The speed and efficiency with which the emergency was handled was directly attributable to the frequency of the emergency preparedness drills carried out onboard the vessel. The inadvertent closing of the port engine room fuel shut-off valve resulted in a blackout of the vessel. The exact cause of the stack fire on No.6 main generator could not be determined. Safety Action Safety Action Taken by the TSB A copy of the TSB Engineering Branch Report LP16/2001 was sent to the generator manufacturer and the owners of the ThebaudSea on 25April2001 for information and remedial action. Safety Action Taken by the Vessel's Owners A complete inspection of all fuel fittings and connections similar to the one that failed was carried out on the five remaining generators onboard the ThebaudSea. No additional problems were detected with any of the other fittings and/or connections. A fleet-wide safety bulletin was issued to all ships' staff, providing guidance on the inspection of the fuel system on the specific make of generators. The fuel oil pressure alarm switches were relocated to eliminate the requirement of running a pressurized fuel oil line in the vicinity of the exhaust manifold. Splash guards were fitted to the fuel oil filters to redirect any potential fuel leaks to the bilge. Electric solenoids on the engine room exhaust ventilation dampers were replaced with a non-electrified manual closing arrangement (pullstation). Safety Action Taken by Transport Canada Transport Canada Marine Safety (TCMS) has previously issued two related Ship Safety Bulletins (SSB) addressing issues similar to the ones involved in this particular occurrence: SSB13/1985 - Fire Precautions with Respect to Fuel Oil Piping and Fittings, issued 18July1985 and SSB08-2000 - Fire Precautions with Respect to Fuel Oil, Lubricating Oil, Hydraulic Oil Piping and Fittings, issued22June2000. These bulletins drew attention to the possibility of fitting shields at connections to contain or deflect oil spray in case of breakage or leakage of piping or fittings and the integrity and condition of fittings. To highlight the potential hazards addressed by the SSBs, TCMS will include the subject of these bulletins as an information item in the November2002 session of the Canadian Marine Advisory Council, Standing Committee on Design and Construction. Safety Action Taken by the Manufacturer To date, TSB is not aware of safety measures instituted by the manufacturer.